OpenC5ISR

OpenLLM

Chat, code agents, and governed multi-model pipelines

Checking Open WebUI
OpenLLM is ready. Choose ChatGPT, Claude, Codex, or the discovered open-source coding lane. The browser never receives provider keys.
Automatic routing

Governed code pipelines

Agents work against an isolated run workspace. Release is blocked at a human gate.

Tests disabled
No workspace files selected.
Codex, Claude Code, and Qwen Code fall back through Open WebUI when their CLI is not installed.

Pipeline runs

Inspect every step, diff, test log, model result, and approval decision.

Select a run.Its execution record will appear here.

Providers and coding agents

Open WebUI is the primary gateway. Direct OpenAI and Anthropic connections remain optional.

Model providers

Coding agent executors

Discovered models

OpenLLM audit stream

Provider calls, pipeline transitions, approvals, artifacts, and OpenBus publication.

Operator token

The token is kept in this browser tab only and sent as a Bearer token for chat and mutation requests.